Using Deep Learning to Refine a Fundamental Stock Selection Model

Summary

The report summary describes a two-stage stock selection process intended to enhance a fundamental factor model. First, the fundamental model creates an initial stock universe; a deep learning model then screens that set again. The stated rationale is to combine the two approaches, with the neural model making a second selection from candidates already identified by fundamental analysis. The source summary claims that the combined approach can deliver greater excess returns while maintaining relatively low turnover. It mentions performance evaluations against the CSI 800 and CSI 300 benchmarks, but the supplied text contains no detailed results, dates, model design, factor definitions, or validation procedure. Because the underlying report is not included, the performance claim cannot be assessed here; the material supports only a high-level description of the proposed workflow and its reported benchmark framing.

Key ideas

• The method applies a deep learning screen after a fundamental model narrows the stock universe.
• The report presents fundamental and deep learning models as complementary selection tools.
• It frames results against the CSI 800 and CSI 300 benchmarks.
• The summary claims improved excess returns at lower turnover but supplies no supporting statistics or validation details.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.